Dear Doctor,
I know how a query can feel. It arrives mid-shift, wrapped in an abundance of misaligned clinical indicators offering a menu of answers that always seems to include the one everyone seems to want. It can feel less like a question and more like forced agreement. I want to be direct about something the query guidelines have stated for years, but that which I don’t think gets communicated clearly enough: you have the freedom and the responsibility to disagree, and we want you to do so. Honestly answering a query sometimes means saying “no.”
Clinical documentation integrity (CDI) is about the accuracy of the health record and how it translates into coded claims data. It is the foundation of compliant billing. Although there is currently much debate in online forums about whether the American Health Information Management Association (AHIMA) and ACDIS have the authority to establish “query compliance,” the Centers for Medicare & Medicaid Services (CMS) is very focused on billing compliance, as the current administration ramps up efforts to reduce fraud. To accurately bill CMS or any payer, the provider must be free to respond honestly, without coercion. To that end, the 2026 AHIMA and ACDIS Guidelines for Achieving a Compliant Query Practice define a compliant query as one that, among other things, “allows the provider to exercise independent clinical judgment.” Inclusion of terms such as “ruled out” and “other explanation of clinical findings, please specify” on multiple-choice queries is not just for appearance. They are “required” options precisely because disagreement, or genuine uncertainty, must remain a real, exercised choice for providers, not a theoretical one nobody actually picks.
I have written in this space about who bears the risk when an artificial intelligence (AI)-generated prompt turns out to be wrong. The same logic runs in the other direction when a human-generated query is what’s applying the pressure. A query is just a question until the provider signs their name to the answer. The moment the provider does, the diagnosis is theirs, not the CDI specialist who asked, and not the hospital that sent it. Recovery Audit Contractor (RAC) auditors, Risk Adjustment Data Validation (RADV) reviewers, and U.S. Department of Justice (DOJ) attorneys subpoena the chart, and the provider’s signature on a query makes them accountable for the associated diagnosis.
Think this is just a hypothetical discussion? It is not. In January 2026, five Kaiser Permanente affiliates agreed to pay $556 million, the largest Medicare Advantage (MA) risk-adjustment settlement in False Claims Act history, to resolve allegations that from 2009 to 2018, the organization pressured physicians to add diagnoses to patient records well after the visits in question. The DOJ’s own complaint described the mechanism in a word CDI professionals will recognize immediately. Kaiser mined patient histories for diagnoses not yet submitted to CMS, then sent “queries” to providers urging them to add those diagnoses through addenda, sometimes more than a year after the encounter. Physicians and facilities that did not meet internal diagnosis-capture targets were singled out, while those who met them received financial bonuses. The second whistleblower, medical coder Ronda Osinek, alleged in the qui tam complaint that Kaiser’s coding practices pressured physicians to add diagnoses.
The government’s complaint named the organization, and the settlement includes no admission of wrongdoing. But the diagnoses at the center of a billion-dollar fraud allegation were sitting in charts that physicians had signed. Dr. James Taylor, a longtime Kaiser physician and one of the whistleblowers, put it this way: physicians are trained to document care truthfully, he said, and “when financial pressure overrides that principle, it undermines trust in the system.” That is not an abstraction. It is what happens when disagreement stops being a real option. The question is whether your organization crossed the tipping point (or is inching toward it).
Kaiser is not an isolated case, and the pattern has not slowed down. In August 2026, a Florida management services organization, Complete Health Partners Holdings, agreed to pay $14.1 million to resolve allegations that its coders identified diagnosis codes tied to substance use disorder and major depressive, bipolar, and paranoid disorder categories and then, in the government’s words, “prompted doctors to add those diagnosis codes, even when the diagnosis codes were unsubstantiated or not clinically justified.” The doctors added them. That single sentence, taken directly from the DOJ’s announcement of the settlement, is the entire argument for why independent judgment cannot be a formality.
The exposure does not stop at the organization named in the press release. Yes, these audits have targeted MA plans, but the next-biggest target is likely short-term acute-care hospitals. Legal analysts tracking this enforcement trend have found that physicians and physician groups whose documentation produced an unsupported diagnosis face their own RADV audit findings, extrapolated repayment demands, and potential False Claims Act liability, particularly where a compensation arrangement rewards a higher risk score. Signing a query is not a neutral act. It creates exposure for the provider, separate from whoever asked the question. Within our industry there is an unwritten expectation that agreement with the most impactful diagnosis is required of providers. This is why we are struggling when it comes to clinical validation queries. Providers have long been told to “exercise their independent judgment” with a wink and a caveat of “as long as it increases reimbursement or improves performance on quality measures.”
Although queries may never reference reimbursement or quality impact, most providers know the desired response. The query guidelines do state that it is noncompliant to keep asking the same question of the same or a different provider when the intent is to pressure, override, or circumvent a provider’s clinical judgment in order to obtain a preferred documentation outcome, but as we introduce more tools to “correct” provider documentation, is that not also creating pressure? Are organizations compliant if the provider first receives a nudge while dictating, followed by a CDI query if they do not acquiesce to the nudge? There is more work to be done to support compliance in how we maintain the integrity of clinical documentation as we introduce AI solutions.
So, Dear Doctor, when a query gives you the option of “unable to determine” or “other explanation of clinical findings,” that is not a trap door hidden in the fine print. It is the point of the exercise. Query professionals and the hospitals that employ them are not doing their job correctly if the goal is your agreement; they are doing their job correctly if the goal is your honest answer, whatever it turns out to be. That protects your license and your reputation considerably more than it protects anyone’s case mix index.
